Image Search System Positional Guidance for Matching Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image recognition systems often provide inadequate search results due to suboptimal image capturing positions, leading to unnecessary processing and increased data traffic.
Innovation Solution
A system and method that directs users to specific image capturing positions relative to an object of interest, utilizing a mobile device and server system to preload reference images with orientation and visual complexity data, guiding users to capture images with higher visual complexity for improved matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If users capture images from arbitrary positions, then image submission is simple, but search result quality deteriorates due to insufficient visual complexity
Solution Approach 1:
The system performs preliminary analysis of reference images to determine optimal capturing positions and visual complexity requirements before the user submits their image. By pre-calculating and storing orientation data and visual complexity metrics for reference images, the system guides users to capture images from positions that will yield better matching results, thus improving search quality without complicating the user workflow
Solution Approach 2:
The system provides feedback to users regarding the quality and orientation of their captured images by comparing them against reference image characteristics. This feedback mechanism enables users to adjust their capturing position or angle to achieve better visual complexity and orientation alignment, thereby improving search result quality while maintaining operational simplicity
2Reliability
If the system processes all submitted images, then comprehensive search coverage is achieved, but data traffic and processing time increase
Solution Approach 1:
The system pre-processes reference images to extract and store orientation data, visual complexity metrics, and key visual features before user submission. This preliminary preparation enables rapid comparison and filtering of user-submitted images against reference images, allowing the system to quickly determine whether an image warrants full processing or can be filtered out, thus reducing unnecessary data traffic and processing while maintaining comprehensive search coverage
Solution Approach 2:
The system performs a preliminary partial processing step by comparing basic image characteristics (orientation, visual complexity) against reference data before committing to full image processing. This two-stage approach processes only the essential filtering criteria first, and only fully processes images that pass the initial screening, thereby reducing overall data traffic and processing energy while maintaining reliable search coverage
3Measurement precision
If multiple reference images with different orientations are stored, then matching accuracy improves, but system complexity increases
Solution Approach 1:
The system segments the reference image database by organizing images into distinct orientation groups or categories. Each reference image is tagged with its orientation metadata, allowing the system to efficiently retrieve and compare only relevant orientation-matched images during search operations. This segmentation reduces the effective search space and simplifies the matching process while maintaining high accuracy through orientation-specific comparisons
Solution Approach 2:
The system adds an orientation dimension to the reference image storage by incorporating orientation metadata and spatial relationship data. This additional dimension enables the system to perform multi-criteria matching (both visual content and orientation), improving matching accuracy. The orientation dimension is implemented through structured data fields rather than physical system changes, thus improving accuracy without proportionally increasing physical system complexity
Data Source
AI summary
A system and method compares an image of an object of interest captured by an image capturing device from a first positional view relative to the object of interest against each of a plurality of images of each of a plurality of reference objects wherein each of the plurality of images of each of the plurality of reference objects is reflective of a unique positional view of the corresponding one of the plurality of reference objects to determine a second positional view relative to the object of interest at which the image capturing device is to be positioned to capture a further image of the product of interest. The further image of the product of interest is then compared against one or more of the plurality of images of one or more of the plurality of reference objects to identify at least one of the plurality of reference objects as being a match for the object of interest whereupon information about the one or more reference objects identified as being a match for the object of interest is provided to a user as a product search result.